UsedTotalWallTime and RequestedTotalWallTime
According to the GLUE 2 specs: https://www.ogf.org/documents/GFD.147.pdf UsedTotalWallTime and RequestedTotalWallTime in the jobs xml file, should both be the sum of each cores walltimes. So for an eight core job that would be walltime-per-core * 8.
I see however that for multicore jobs the RequestedTotalWallTime is given per core, while the UsedTotalWallTime is summed up for all cores (the latter according to the GLUE2 specs). An example showing one running job: job.yuRLDmrWCDvnKwAJwmkId7EnABFKDmABFKDm1AMKDmLiPKDmivL88n.xml: 51840 job.yuRLDmrWCDvnKwAJwmkId7EnABFKDmABFKDm1AMKDmLiPKDmivL88n.xml: 156240
Comparing with the walltime from this jobs jdl description from Panda (corresponding to the requestedtotalwalltime) this job has (walltime=864)(cputime=6912) - in minutes. So walltime is 864*60=51840 s which is the number in the xml file.
So to me it seems like the requestedtotalwalltime is not multiplied by number of cores, but the usedtotalwalltime is.
The job shown as an example is this one: https://aipanda403.cern.ch/data/jobs/2019-07-31/UIO_CLOUD/4434164910.jdl And its full xml file is
[root@frontend001 control]# cat job.yuRLDmrWCDvnKwAJwmkId7EnABFKDmABFKDm1AMKDmLiPKDmivL88n.xml
<ComputingActivity xmlns="http://schemas.ogf.org/glue/2009/03/spec_2.0_r1" BaseType="Activity" CreationTime="2019-08-01T06:56:49Z" Validity="10800">
<ID>urn:caid:frontend001.grid.uiocloud.no:org.ogf.glue.emies.activitycreation:yuRLDmrWCDvnKwAJwmkId7EnABFKDmABFKDm1AMKDmLiPKDmivL88n</ID>
<Name>mc16_13TeV.410472.PhPy8EG_A14_ttbar_hdamp258p75_dil.deriv.e6348_e5984_s3126_r10724_r10726_p3931.4433549934</Name>
<OtherInfo>SubmittedVia=org.ogf.glue.emies.activitycreation</OtherInfo>
<Type>single</Type>
<IDFromEndpoint>urn:idfe:yuRLDmrWCDvnKwAJwmkId7EnABFKDmABFKDm1AMKDmLiPKDmivL88n</IDFromEndpoint>
<LocalIDFromManager>59990</LocalIDFromManager>
<JobDescription>emies:adl</JobDescription>
<State>nordugrid:INLRMS:R</State>
<State>file:inlrms</State>
<State>bes:Running</State>
<State>emies:processing-running</State>
<State>emiesattr:app-running</State>
<WaitingPosition>0</WaitingPosition>
<Owner>/DC=ch/DC=cern/OU=Organic Units/OU=Users/CN=atlact1/CN=555105/CN=Robot: ATLAS aCT 1</Owner>
<LocalOwner>grid</LocalOwner>
<RequestedTotalWallTime>51840</RequestedTotalWallTime>
<RequestedTotalCPUTime>414720</RequestedTotalCPUTime>
<RequestedSlots>8</RequestedSlots>
<RequestedApplicationEnvironment>ENV/PROXY</RequestedApplicationEnvironment>
<StdIn>/dev/null</StdIn>
<StdOut>log.18750243._000382.job.log.1</StdOut>
<StdErr>log.18750243._000382.job.log.1</StdErr>
<LogDir>gmlog</LogDir>
<ExecutionNode>compute001</ExecutionNode>
<Queue>main</Queue>
<UsedTotalWallTime>159696</UsedTotalWallTime>
<UsedTotalCPUTime>159696</UsedTotalCPUTime>
<UsedMainMemory>29</UsedMainMemory>
<SubmissionTime>2019-07-31T22:45:33Z</SubmissionTime>
<ProxyExpirationTime>2019-08-04T12:43:26Z</ProxyExpirationTime>
<SubmissionHost>188.185.76.169</SubmissionHost>
<OtherMessages>mc16_13TeV_4104</OtherMessages>
<Associations>
<ComputingShareID>urn:ogf:ComputingShare:frontend001.grid.uiocloud.no:main</ComputingShareID>
</Associations>
</ComputingActivity>